The idea is to have a separate instance in a different physical machine serving the same trading pairs as the first one. The failover instance will not receive orders directly from users but sync trades and balances from first instance, if the first instance fails, then this instance will take over the traffic automatically.
The idea is to have a separate instance in a different physical machine serving the same trading pairs as the first one. The failover instance will not receive orders directly from users but sync trades and balances from first instance, if the first instance fails, then this instance will take over the traffic automatically.